1 // Copyright 2009 The Go Authors. All rights reserved.
2 // Use of this source code is governed by a BSD-style
3 // license that can be found in the LICENSE file.
5 // Garbage collector (GC).
7 // The GC runs concurrently with mutator threads, is type accurate (aka precise), allows multiple
8 // GC thread to run in parallel. It is a concurrent mark and sweep that uses a write barrier. It is
9 // non-generational and non-compacting. Allocation is done using size segregated per P allocation
10 // areas to minimize fragmentation while eliminating locks in the common case.
12 // The algorithm decomposes into several steps.
13 // This is a high level description of the algorithm being used. For an overview of GC a good
14 // place to start is Richard Jones' gchandbook.org.
16 // The algorithm's intellectual heritage includes Dijkstra's on-the-fly algorithm, see
17 // Edsger W. Dijkstra, Leslie Lamport, A. J. Martin, C. S. Scholten, and E. F. M. Steffens. 1978.
18 // On-the-fly garbage collection: an exercise in cooperation. Commun. ACM 21, 11 (November 1978),
20 // For journal quality proofs that these steps are complete, correct, and terminate see
21 // Hudson, R., and Moss, J.E.B. Copying Garbage Collection without stopping the world.
22 // Concurrency and Computation: Practice and Experience 15(3-5), 2003.
24 // 1. GC performs sweep termination.
26 // a. Stop the world. This causes all Ps to reach a GC safe-point.
28 // b. Sweep any unswept spans. There will only be unswept spans if
29 // this GC cycle was forced before the expected time.
31 // 2. GC performs the mark phase.
33 // a. Prepare for the mark phase by setting gcphase to _GCmark
34 // (from _GCoff), enabling the write barrier, enabling mutator
35 // assists, and enqueueing root mark jobs. No objects may be
36 // scanned until all Ps have enabled the write barrier, which is
37 // accomplished using STW.
39 // b. Start the world. From this point, GC work is done by mark
40 // workers started by the scheduler and by assists performed as
41 // part of allocation. The write barrier shades both the
42 // overwritten pointer and the new pointer value for any pointer
43 // writes (see mbarrier.go for details). Newly allocated objects
44 // are immediately marked black.
46 // c. GC performs root marking jobs. This includes scanning all
47 // stacks, shading all globals, and shading any heap pointers in
48 // off-heap runtime data structures. Scanning a stack stops a
49 // goroutine, shades any pointers found on its stack, and then
50 // resumes the goroutine.
52 // d. GC drains the work queue of grey objects, scanning each grey
53 // object to black and shading all pointers found in the object
54 // (which in turn may add those pointers to the work queue).
56 // e. Because GC work is spread across local caches, GC uses a
57 // distributed termination algorithm to detect when there are no
58 // more root marking jobs or grey objects (see gcMarkDone). At this
59 // point, GC transitions to mark termination.
61 // 3. GC performs mark termination.
65 // b. Set gcphase to _GCmarktermination, and disable workers and
68 // c. Perform housekeeping like flushing mcaches.
70 // 4. GC performs the sweep phase.
72 // a. Prepare for the sweep phase by setting gcphase to _GCoff,
73 // setting up sweep state and disabling the write barrier.
75 // b. Start the world. From this point on, newly allocated objects
76 // are white, and allocating sweeps spans before use if necessary.
78 // c. GC does concurrent sweeping in the background and in response
79 // to allocation. See description below.
81 // 5. When sufficient allocation has taken place, replay the sequence
82 // starting with 1 above. See discussion of GC rate below.
86 // The sweep phase proceeds concurrently with normal program execution.
87 // The heap is swept span-by-span both lazily (when a goroutine needs another span)
88 // and concurrently in a background goroutine (this helps programs that are not CPU bound).
89 // At the end of STW mark termination all spans are marked as "needs sweeping".
91 // The background sweeper goroutine simply sweeps spans one-by-one.
93 // To avoid requesting more OS memory while there are unswept spans, when a
94 // goroutine needs another span, it first attempts to reclaim that much memory
95 // by sweeping. When a goroutine needs to allocate a new small-object span, it
96 // sweeps small-object spans for the same object size until it frees at least
97 // one object. When a goroutine needs to allocate large-object span from heap,
98 // it sweeps spans until it frees at least that many pages into heap. There is
99 // one case where this may not suffice: if a goroutine sweeps and frees two
100 // nonadjacent one-page spans to the heap, it will allocate a new two-page
101 // span, but there can still be other one-page unswept spans which could be
102 // combined into a two-page span.
104 // It's critical to ensure that no operations proceed on unswept spans (that would corrupt
105 // mark bits in GC bitmap). During GC all mcaches are flushed into the central cache,
106 // so they are empty. When a goroutine grabs a new span into mcache, it sweeps it.
107 // When a goroutine explicitly frees an object or sets a finalizer, it ensures that
108 // the span is swept (either by sweeping it, or by waiting for the concurrent sweep to finish).
109 // The finalizer goroutine is kicked off only when all spans are swept.
110 // When the next GC starts, it sweeps all not-yet-swept spans (if any).
113 // Next GC is after we've allocated an extra amount of memory proportional to
114 // the amount already in use. The proportion is controlled by GOGC environment variable
115 // (100 by default). If GOGC=100 and we're using 4M, we'll GC again when we get to 8M
116 // (this mark is computed by the gcController.heapGoal method). This keeps the GC cost in
117 // linear proportion to the allocation cost. Adjusting GOGC just changes the linear constant
118 // (and also the amount of extra memory used).
122 // In order to prevent long pauses while scanning large objects and to
123 // improve parallelism, the garbage collector breaks up scan jobs for
124 // objects larger than maxObletBytes into "oblets" of at most
125 // maxObletBytes. When scanning encounters the beginning of a large
126 // object, it scans only the first oblet and enqueues the remaining
127 // oblets as new scan jobs.
133 "runtime/internal/atomic"
139 _FinBlockSize = 4 * 1024
141 // concurrentSweep is a debug flag. Disabling this flag
142 // ensures all spans are swept while the world is stopped.
143 concurrentSweep = true
145 // debugScanConservative enables debug logging for stack
146 // frames that are scanned conservatively.
147 debugScanConservative = false
149 // sweepMinHeapDistance is a lower bound on the heap distance
150 // (in bytes) reserved for concurrent sweeping between GC
152 sweepMinHeapDistance = 1024 * 1024
155 // heapObjectsCanMove always returns false in the current garbage collector.
156 // It exists for go4.org/unsafe/assume-no-moving-gc, which is an
157 // unfortunate idea that had an even more unfortunate implementation.
158 // Every time a new Go release happened, the package stopped building,
159 // and the authors had to add a new file with a new //go:build line, and
160 // then the entire ecosystem of packages with that as a dependency had to
161 // explicitly update to the new version. Many packages depend on
162 // assume-no-moving-gc transitively, through paths like
163 // inet.af/netaddr -> go4.org/intern -> assume-no-moving-gc.
164 // This was causing a significant amount of friction around each new
165 // release, so we added this bool for the package to //go:linkname
166 // instead. The bool is still unfortunate, but it's not as bad as
167 // breaking the ecosystem on every new release.
169 // If the Go garbage collector ever does move heap objects, we can set
170 // this to true to break all the programs using assume-no-moving-gc.
172 //go:linkname heapObjectsCanMove
173 func heapObjectsCanMove() bool {
178 if unsafe.Sizeof(workbuf{}) != _WorkbufSize {
179 throw("size of Workbuf is suboptimal")
181 // No sweep on the first cycle.
182 sweep.active.state.Store(sweepDrainedMask)
184 // Initialize GC pacer state.
185 // Use the environment variable GOGC for the initial gcPercent value.
186 // Use the environment variable GOMEMLIMIT for the initial memoryLimit value.
187 gcController.init(readGOGC(), readGOMEMLIMIT())
190 work.markDoneSema = 1
191 lockInit(&work.sweepWaiters.lock, lockRankSweepWaiters)
192 lockInit(&work.assistQueue.lock, lockRankAssistQueue)
193 lockInit(&work.wbufSpans.lock, lockRankWbufSpans)
196 // gcenable is called after the bulk of the runtime initialization,
197 // just before we're about to start letting user code run.
198 // It kicks off the background sweeper goroutine, the background
199 // scavenger goroutine, and enables GC.
201 // Kick off sweeping and scavenging.
202 c := make(chan int, 2)
207 memstats.enablegc = true // now that runtime is initialized, GC is okay
210 // Garbage collector phase.
211 // Indicates to write barrier and synchronization task to perform.
214 // The compiler knows about this variable.
215 // If you change it, you must change builtin/runtime.go, too.
216 // If you change the first four bytes, you must also change the write
217 // barrier insertion code.
218 var writeBarrier struct {
219 enabled bool // compiler emits a check of this before calling write barrier
220 pad [3]byte // compiler uses 32-bit load for "enabled" field
221 alignme uint64 // guarantee alignment so that compiler can use a 32 or 64-bit load
224 // gcBlackenEnabled is 1 if mutator assists and background mark
225 // workers are allowed to blacken objects. This must only be set when
226 // gcphase == _GCmark.
227 var gcBlackenEnabled uint32
230 _GCoff = iota // GC not running; sweeping in background, write barrier disabled
231 _GCmark // GC marking roots and workbufs: allocate black, write barrier ENABLED
232 _GCmarktermination // GC mark termination: allocate black, P's help GC, write barrier ENABLED
236 func setGCPhase(x uint32) {
237 atomic.Store(&gcphase, x)
238 writeBarrier.enabled = gcphase == _GCmark || gcphase == _GCmarktermination
241 // gcMarkWorkerMode represents the mode that a concurrent mark worker
242 // should operate in.
244 // Concurrent marking happens through four different mechanisms. One
245 // is mutator assists, which happen in response to allocations and are
246 // not scheduled. The other three are variations in the per-P mark
247 // workers and are distinguished by gcMarkWorkerMode.
248 type gcMarkWorkerMode int
251 // gcMarkWorkerNotWorker indicates that the next scheduled G is not
252 // starting work and the mode should be ignored.
253 gcMarkWorkerNotWorker gcMarkWorkerMode = iota
255 // gcMarkWorkerDedicatedMode indicates that the P of a mark
256 // worker is dedicated to running that mark worker. The mark
257 // worker should run without preemption.
258 gcMarkWorkerDedicatedMode
260 // gcMarkWorkerFractionalMode indicates that a P is currently
261 // running the "fractional" mark worker. The fractional worker
262 // is necessary when GOMAXPROCS*gcBackgroundUtilization is not
263 // an integer and using only dedicated workers would result in
264 // utilization too far from the target of gcBackgroundUtilization.
265 // The fractional worker should run until it is preempted and
266 // will be scheduled to pick up the fractional part of
267 // GOMAXPROCS*gcBackgroundUtilization.
268 gcMarkWorkerFractionalMode
270 // gcMarkWorkerIdleMode indicates that a P is running the mark
271 // worker because it has nothing else to do. The idle worker
272 // should run until it is preempted and account its time
273 // against gcController.idleMarkTime.
277 // gcMarkWorkerModeStrings are the strings labels of gcMarkWorkerModes
278 // to use in execution traces.
279 var gcMarkWorkerModeStrings = [...]string{
286 // pollFractionalWorkerExit reports whether a fractional mark worker
287 // should self-preempt. It assumes it is called from the fractional
289 func pollFractionalWorkerExit() bool {
290 // This should be kept in sync with the fractional worker
291 // scheduler logic in findRunnableGCWorker.
293 delta := now - gcController.markStartTime
297 p := getg().m.p.ptr()
298 selfTime := p.gcFractionalMarkTime + (now - p.gcMarkWorkerStartTime)
299 // Add some slack to the utilization goal so that the
300 // fractional worker isn't behind again the instant it exits.
301 return float64(selfTime)/float64(delta) > 1.2*gcController.fractionalUtilizationGoal
306 type workType struct {
307 full lfstack // lock-free list of full blocks workbuf
308 _ cpu.CacheLinePad // prevents false-sharing between full and empty
309 empty lfstack // lock-free list of empty blocks workbuf
310 _ cpu.CacheLinePad // prevents false-sharing between empty and nproc/nwait
314 // free is a list of spans dedicated to workbufs, but
315 // that don't currently contain any workbufs.
317 // busy is a list of all spans containing workbufs on
318 // one of the workbuf lists.
322 // Restore 64-bit alignment on 32-bit.
325 // bytesMarked is the number of bytes marked this cycle. This
326 // includes bytes blackened in scanned objects, noscan objects
327 // that go straight to black, and permagrey objects scanned by
328 // markroot during the concurrent scan phase. This is updated
329 // atomically during the cycle. Updates may be batched
330 // arbitrarily, since the value is only read at the end of the
333 // Because of benign races during marking, this number may not
334 // be the exact number of marked bytes, but it should be very
337 // Put this field here because it needs 64-bit atomic access
338 // (and thus 8-byte alignment even on 32-bit architectures).
341 markrootNext uint32 // next markroot job
342 markrootJobs uint32 // number of markroot jobs
348 // Number of roots of various root types. Set by gcMarkRootPrepare.
350 // nStackRoots == len(stackRoots), but we have nStackRoots for
352 nDataRoots, nBSSRoots, nSpanRoots, nStackRoots int
354 // Base indexes of each root type. Set by gcMarkRootPrepare.
355 baseData, baseBSS, baseSpans, baseStacks, baseEnd uint32
357 // stackRoots is a snapshot of all of the Gs that existed
358 // before the beginning of concurrent marking. The backing
359 // store of this must not be modified because it might be
360 // shared with allgs.
363 // Each type of GC state transition is protected by a lock.
364 // Since multiple threads can simultaneously detect the state
365 // transition condition, any thread that detects a transition
366 // condition must acquire the appropriate transition lock,
367 // re-check the transition condition and return if it no
368 // longer holds or perform the transition if it does.
369 // Likewise, any transition must invalidate the transition
370 // condition before releasing the lock. This ensures that each
371 // transition is performed by exactly one thread and threads
372 // that need the transition to happen block until it has
375 // startSema protects the transition from "off" to mark or
378 // markDoneSema protects transitions from mark to mark termination.
381 bgMarkReady note // signal background mark worker has started
382 bgMarkDone uint32 // cas to 1 when at a background mark completion point
383 // Background mark completion signaling
385 // mode is the concurrency mode of the current GC cycle.
388 // userForced indicates the current GC cycle was forced by an
389 // explicit user call.
392 // initialHeapLive is the value of gcController.heapLive at the
393 // beginning of this GC cycle.
394 initialHeapLive uint64
396 // assistQueue is a queue of assists that are blocked because
397 // there was neither enough credit to steal or enough work to
404 // sweepWaiters is a list of blocked goroutines to wake when
405 // we transition from mark termination to sweep.
406 sweepWaiters struct {
411 // cycles is the number of completed GC cycles, where a GC
412 // cycle is sweep termination, mark, mark termination, and
413 // sweep. This differs from memstats.numgc, which is
414 // incremented at mark termination.
417 // Timing/utilization stats for this cycle.
418 stwprocs, maxprocs int32
419 tSweepTerm, tMark, tMarkTerm, tEnd int64 // nanotime() of phase start
421 pauseNS int64 // total STW time this cycle
422 pauseStart int64 // nanotime() of last STW
424 // debug.gctrace heap sizes for this cycle.
425 heap0, heap1, heap2 uint64
427 // Cumulative estimated CPU usage.
431 // GC runs a garbage collection and blocks the caller until the
432 // garbage collection is complete. It may also block the entire
435 // We consider a cycle to be: sweep termination, mark, mark
436 // termination, and sweep. This function shouldn't return
437 // until a full cycle has been completed, from beginning to
438 // end. Hence, we always want to finish up the current cycle
439 // and start a new one. That means:
441 // 1. In sweep termination, mark, or mark termination of cycle
442 // N, wait until mark termination N completes and transitions
445 // 2. In sweep N, help with sweep N.
447 // At this point we can begin a full cycle N+1.
449 // 3. Trigger cycle N+1 by starting sweep termination N+1.
451 // 4. Wait for mark termination N+1 to complete.
453 // 5. Help with sweep N+1 until it's done.
455 // This all has to be written to deal with the fact that the
456 // GC may move ahead on its own. For example, when we block
457 // until mark termination N, we may wake up in cycle N+2.
459 // Wait until the current sweep termination, mark, and mark
460 // termination complete.
461 n := work.cycles.Load()
464 // We're now in sweep N or later. Trigger GC cycle N+1, which
465 // will first finish sweep N if necessary and then enter sweep
467 gcStart(gcTrigger{kind: gcTriggerCycle, n: n + 1})
469 // Wait for mark termination N+1 to complete.
472 // Finish sweep N+1 before returning. We do this both to
473 // complete the cycle and because runtime.GC() is often used
474 // as part of tests and benchmarks to get the system into a
475 // relatively stable and isolated state.
476 for work.cycles.Load() == n+1 && sweepone() != ^uintptr(0) {
480 // Callers may assume that the heap profile reflects the
481 // just-completed cycle when this returns (historically this
482 // happened because this was a STW GC), but right now the
483 // profile still reflects mark termination N, not N+1.
485 // As soon as all of the sweep frees from cycle N+1 are done,
486 // we can go ahead and publish the heap profile.
488 // First, wait for sweeping to finish. (We know there are no
489 // more spans on the sweep queue, but we may be concurrently
490 // sweeping spans, so we have to wait.)
491 for work.cycles.Load() == n+1 && !isSweepDone() {
495 // Now we're really done with sweeping, so we can publish the
496 // stable heap profile. Only do this if we haven't already hit
497 // another mark termination.
499 cycle := work.cycles.Load()
500 if cycle == n+1 || (gcphase == _GCmark && cycle == n+2) {
506 // gcWaitOnMark blocks until GC finishes the Nth mark phase. If GC has
507 // already completed this mark phase, it returns immediately.
508 func gcWaitOnMark(n uint32) {
510 // Disable phase transitions.
511 lock(&work.sweepWaiters.lock)
512 nMarks := work.cycles.Load()
513 if gcphase != _GCmark {
514 // We've already completed this cycle's mark.
519 unlock(&work.sweepWaiters.lock)
523 // Wait until sweep termination, mark, and mark
524 // termination of cycle N complete.
525 work.sweepWaiters.list.push(getg())
526 goparkunlock(&work.sweepWaiters.lock, waitReasonWaitForGCCycle, traceBlockUntilGCEnds, 1)
530 // gcMode indicates how concurrent a GC cycle should be.
534 gcBackgroundMode gcMode = iota // concurrent GC and sweep
535 gcForceMode // stop-the-world GC now, concurrent sweep
536 gcForceBlockMode // stop-the-world GC now and STW sweep (forced by user)
539 // A gcTrigger is a predicate for starting a GC cycle. Specifically,
540 // it is an exit condition for the _GCoff phase.
541 type gcTrigger struct {
543 now int64 // gcTriggerTime: current time
544 n uint32 // gcTriggerCycle: cycle number to start
547 type gcTriggerKind int
550 // gcTriggerHeap indicates that a cycle should be started when
551 // the heap size reaches the trigger heap size computed by the
553 gcTriggerHeap gcTriggerKind = iota
555 // gcTriggerTime indicates that a cycle should be started when
556 // it's been more than forcegcperiod nanoseconds since the
557 // previous GC cycle.
560 // gcTriggerCycle indicates that a cycle should be started if
561 // we have not yet started cycle number gcTrigger.n (relative
566 // test reports whether the trigger condition is satisfied, meaning
567 // that the exit condition for the _GCoff phase has been met. The exit
568 // condition should be tested when allocating.
569 func (t gcTrigger) test() bool {
570 if !memstats.enablegc || panicking.Load() != 0 || gcphase != _GCoff {
575 trigger, _ := gcController.trigger()
576 return gcController.heapLive.Load() >= trigger
578 if gcController.gcPercent.Load() < 0 {
581 lastgc := int64(atomic.Load64(&memstats.last_gc_nanotime))
582 return lastgc != 0 && t.now-lastgc > forcegcperiod
584 // t.n > work.cycles, but accounting for wraparound.
585 return int32(t.n-work.cycles.Load()) > 0
590 // gcStart starts the GC. It transitions from _GCoff to _GCmark (if
591 // debug.gcstoptheworld == 0) or performs all of GC (if
592 // debug.gcstoptheworld != 0).
594 // This may return without performing this transition in some cases,
595 // such as when called on a system stack or with locks held.
596 func gcStart(trigger gcTrigger) {
597 // Since this is called from malloc and malloc is called in
598 // the guts of a number of libraries that might be holding
599 // locks, don't attempt to start GC in non-preemptible or
600 // potentially unstable situations.
602 if gp := getg(); gp == mp.g0 || mp.locks > 1 || mp.preemptoff != "" {
609 // Pick up the remaining unswept/not being swept spans concurrently
611 // This shouldn't happen if we're being invoked in background
612 // mode since proportional sweep should have just finished
613 // sweeping everything, but rounding errors, etc, may leave a
614 // few spans unswept. In forced mode, this is necessary since
615 // GC can be forced at any point in the sweeping cycle.
617 // We check the transition condition continuously here in case
618 // this G gets delayed in to the next GC cycle.
619 for trigger.test() && sweepone() != ^uintptr(0) {
622 // Perform GC initialization and the sweep termination
624 semacquire(&work.startSema)
625 // Re-check transition condition under transition lock.
627 semrelease(&work.startSema)
631 // In gcstoptheworld debug mode, upgrade the mode accordingly.
632 // We do this after re-checking the transition condition so
633 // that multiple goroutines that detect the heap trigger don't
634 // start multiple STW GCs.
635 mode := gcBackgroundMode
636 if debug.gcstoptheworld == 1 {
638 } else if debug.gcstoptheworld == 2 {
639 mode = gcForceBlockMode
642 // Ok, we're doing it! Stop everybody else
644 semacquire(&worldsema)
646 // For stats, check if this GC was forced by the user.
647 // Update it under gcsema to avoid gctrace getting wrong values.
648 work.userForced = trigger.kind == gcTriggerCycle
650 trace := traceAcquire()
656 // Check that all Ps have finished deferred mcache flushes.
657 for _, p := range allp {
658 if fg := p.mcache.flushGen.Load(); fg != mheap_.sweepgen {
659 println("runtime: p", p.id, "flushGen", fg, "!= sweepgen", mheap_.sweepgen)
660 throw("p mcache not flushed")
664 gcBgMarkStartWorkers()
666 systemstack(gcResetMarkState)
668 work.stwprocs, work.maxprocs = gomaxprocs, gomaxprocs
669 if work.stwprocs > ncpu {
670 // This is used to compute CPU time of the STW phases,
671 // so it can't be more than ncpu, even if GOMAXPROCS is.
674 work.heap0 = gcController.heapLive.Load()
679 work.tSweepTerm = now
680 work.pauseStart = now
681 systemstack(func() { stopTheWorldWithSema(stwGCSweepTerm) })
682 // Finish sweep before we start concurrent scan.
687 // clearpools before we start the GC. If we wait the memory will not be
688 // reclaimed until the next GC cycle.
693 // Assists and workers can start the moment we start
695 gcController.startCycle(now, int(gomaxprocs), trigger)
697 // Notify the CPU limiter that assists may begin.
698 gcCPULimiter.startGCTransition(true, now)
700 // In STW mode, disable scheduling of user Gs. This may also
701 // disable scheduling of this goroutine, so it may block as
702 // soon as we start the world again.
703 if mode != gcBackgroundMode {
704 schedEnableUser(false)
707 // Enter concurrent mark phase and enable
710 // Because the world is stopped, all Ps will
711 // observe that write barriers are enabled by
712 // the time we start the world and begin
715 // Write barriers must be enabled before assists are
716 // enabled because they must be enabled before
717 // any non-leaf heap objects are marked. Since
718 // allocations are blocked until assists can
719 // happen, we want to enable assists as early as
723 gcBgMarkPrepare() // Must happen before assists are enabled.
726 // Mark all active tinyalloc blocks. Since we're
727 // allocating from these, they need to be black like
728 // other allocations. The alternative is to blacken
729 // the tiny block on every allocation from it, which
730 // would slow down the tiny allocator.
733 // At this point all Ps have enabled the write
734 // barrier, thus maintaining the no white to
735 // black invariant. Enable mutator assists to
736 // put back-pressure on fast allocating
738 atomic.Store(&gcBlackenEnabled, 1)
740 // In STW mode, we could block the instant systemstack
741 // returns, so make sure we're not preemptible.
746 now = startTheWorldWithSema()
747 work.pauseNS += now - work.pauseStart
749 memstats.gcPauseDist.record(now - work.pauseStart)
751 sweepTermCpu := int64(work.stwprocs) * (work.tMark - work.tSweepTerm)
752 work.cpuStats.gcPauseTime += sweepTermCpu
753 work.cpuStats.gcTotalTime += sweepTermCpu
755 // Release the CPU limiter.
756 gcCPULimiter.finishGCTransition(now)
759 // Release the world sema before Gosched() in STW mode
760 // because we will need to reacquire it later but before
761 // this goroutine becomes runnable again, and we could
762 // self-deadlock otherwise.
763 semrelease(&worldsema)
766 // Make sure we block instead of returning to user code
768 if mode != gcBackgroundMode {
772 semrelease(&work.startSema)
775 // gcMarkDoneFlushed counts the number of P's with flushed work.
777 // Ideally this would be a captured local in gcMarkDone, but forEachP
778 // escapes its callback closure, so it can't capture anything.
780 // This is protected by markDoneSema.
781 var gcMarkDoneFlushed uint32
783 // gcMarkDone transitions the GC from mark to mark termination if all
784 // reachable objects have been marked (that is, there are no grey
785 // objects and can be no more in the future). Otherwise, it flushes
786 // all local work to the global queues where it can be discovered by
789 // This should be called when all local mark work has been drained and
790 // there are no remaining workers. Specifically, when
792 // work.nwait == work.nproc && !gcMarkWorkAvailable(p)
794 // The calling context must be preemptible.
796 // Flushing local work is important because idle Ps may have local
797 // work queued. This is the only way to make that work visible and
798 // drive GC to completion.
800 // It is explicitly okay to have write barriers in this function. If
801 // it does transition to mark termination, then all reachable objects
802 // have been marked, so the write barrier cannot shade any more
805 // Ensure only one thread is running the ragged barrier at a
807 semacquire(&work.markDoneSema)
810 // Re-check transition condition under transition lock.
812 // It's critical that this checks the global work queues are
813 // empty before performing the ragged barrier. Otherwise,
814 // there could be global work that a P could take after the P
815 // has passed the ragged barrier.
816 if !(gcphase == _GCmark && work.nwait == work.nproc && !gcMarkWorkAvailable(nil)) {
817 semrelease(&work.markDoneSema)
821 // forEachP needs worldsema to execute, and we'll need it to
822 // stop the world later, so acquire worldsema now.
823 semacquire(&worldsema)
825 // Flush all local buffers and collect flushedWork flags.
826 gcMarkDoneFlushed = 0
827 forEachP(waitReasonGCMarkTermination, func(pp *p) {
828 // Flush the write barrier buffer, since this may add
829 // work to the gcWork.
832 // Flush the gcWork, since this may create global work
833 // and set the flushedWork flag.
835 // TODO(austin): Break up these workbufs to
836 // better distribute work.
838 // Collect the flushedWork flag.
839 if pp.gcw.flushedWork {
840 atomic.Xadd(&gcMarkDoneFlushed, 1)
841 pp.gcw.flushedWork = false
845 if gcMarkDoneFlushed != 0 {
846 // More grey objects were discovered since the
847 // previous termination check, so there may be more
848 // work to do. Keep going. It's possible the
849 // transition condition became true again during the
850 // ragged barrier, so re-check it.
851 semrelease(&worldsema)
855 // There was no global work, no local work, and no Ps
856 // communicated work since we took markDoneSema. Therefore
857 // there are no grey objects and no more objects can be
858 // shaded. Transition to mark termination.
861 work.pauseStart = now
862 getg().m.preemptoff = "gcing"
863 systemstack(func() { stopTheWorldWithSema(stwGCMarkTerm) })
864 // The gcphase is _GCmark, it will transition to _GCmarktermination
865 // below. The important thing is that the wb remains active until
866 // all marking is complete. This includes writes made by the GC.
868 // There is sometimes work left over when we enter mark termination due
869 // to write barriers performed after the completion barrier above.
870 // Detect this and resume concurrent mark. This is obviously
873 // See issue #27993 for details.
875 // Switch to the system stack to call wbBufFlush1, though in this case
876 // it doesn't matter because we're non-preemptible anyway.
879 for _, p := range allp {
888 getg().m.preemptoff = ""
890 now := startTheWorldWithSema()
891 work.pauseNS += now - work.pauseStart
892 memstats.gcPauseDist.record(now - work.pauseStart)
894 semrelease(&worldsema)
898 gcComputeStartingStackSize()
900 // Disable assists and background workers. We must do
901 // this before waking blocked assists.
902 atomic.Store(&gcBlackenEnabled, 0)
904 // Notify the CPU limiter that GC assists will now cease.
905 gcCPULimiter.startGCTransition(false, now)
907 // Wake all blocked assists. These will run when we
908 // start the world again.
911 // Likewise, release the transition lock. Blocked
912 // workers and assists will run when we start the
914 semrelease(&work.markDoneSema)
916 // In STW mode, re-enable user goroutines. These will be
917 // queued to run after we start the world.
918 schedEnableUser(true)
920 // endCycle depends on all gcWork cache stats being flushed.
921 // The termination algorithm above ensured that up to
922 // allocations since the ragged barrier.
923 gcController.endCycle(now, int(gomaxprocs), work.userForced)
925 // Perform mark termination. This will restart the world.
929 // World must be stopped and mark assists and background workers must be
931 func gcMarkTermination() {
932 // Start marktermination (write barrier remains enabled for now).
933 setGCPhase(_GCmarktermination)
935 work.heap1 = gcController.heapLive.Load()
936 startTime := nanotime()
939 mp.preemptoff = "gcing"
942 casGToWaiting(curgp, _Grunning, waitReasonGarbageCollection)
944 // Run gc on the g0 stack. We do this so that the g stack
945 // we're currently running on will no longer change. Cuts
946 // the root set down a bit (g0 stacks are not scanned, and
947 // we don't need to scan gc's internal state). We also
948 // need to switch to g0 so we can shrink the stack.
951 // Must return immediately.
952 // The outer function's stack may have moved
953 // during gcMark (it shrinks stacks, including the
954 // outer function's stack), so we must not refer
955 // to any of its variables. Return back to the
956 // non-system stack to pick up the new addresses
957 // before continuing.
962 work.heap2 = work.bytesMarked
963 if debug.gccheckmark > 0 {
964 // Run a full non-parallel, stop-the-world
965 // mark using checkmark bits, to check that we
966 // didn't forget to mark anything during the
967 // concurrent mark process.
970 gcw := &getg().m.p.ptr().gcw
972 wbBufFlush1(getg().m.p.ptr())
977 // marking is complete so we can turn the write barrier off
979 stwSwept = gcSweep(work.mode)
983 casgstatus(curgp, _Gwaiting, _Grunning)
985 trace := traceAcquire()
994 if gcphase != _GCoff {
995 throw("gc done but gcphase != _GCoff")
998 // Record heapInUse for scavenger.
999 memstats.lastHeapInUse = gcController.heapInUse.load()
1001 // Update GC trigger and pacing, as well as downstream consumers
1002 // of this pacing information, for the next cycle.
1003 systemstack(gcControllerCommit)
1005 // Update timing memstats
1007 sec, nsec, _ := time_now()
1008 unixNow := sec*1e9 + int64(nsec)
1009 work.pauseNS += now - work.pauseStart
1011 memstats.gcPauseDist.record(now - work.pauseStart)
1012 atomic.Store64(&memstats.last_gc_unix, uint64(unixNow)) // must be Unix time to make sense to user
1013 atomic.Store64(&memstats.last_gc_nanotime, uint64(now)) // monotonic time for us
1014 memstats.pause_ns[memstats.numgc%uint32(len(memstats.pause_ns))] = uint64(work.pauseNS)
1015 memstats.pause_end[memstats.numgc%uint32(len(memstats.pause_end))] = uint64(unixNow)
1016 memstats.pause_total_ns += uint64(work.pauseNS)
1018 markTermCpu := int64(work.stwprocs) * (work.tEnd - work.tMarkTerm)
1019 work.cpuStats.gcPauseTime += markTermCpu
1020 work.cpuStats.gcTotalTime += markTermCpu
1022 // Accumulate CPU stats.
1024 // Pass gcMarkPhase=true so we can get all the latest GC CPU stats in there too.
1025 work.cpuStats.accumulate(now, true)
1027 // Compute overall GC CPU utilization.
1028 // Omit idle marking time from the overall utilization here since it's "free".
1029 memstats.gc_cpu_fraction = float64(work.cpuStats.gcTotalTime-work.cpuStats.gcIdleTime) / float64(work.cpuStats.totalTime)
1031 // Reset assist time and background time stats.
1033 // Do this now, instead of at the start of the next GC cycle, because
1034 // these two may keep accumulating even if the GC is not active.
1035 scavenge.assistTime.Store(0)
1036 scavenge.backgroundTime.Store(0)
1038 // Reset idle time stat.
1039 sched.idleTime.Store(0)
1041 if work.userForced {
1042 memstats.numforcedgc++
1045 // Bump GC cycle count and wake goroutines waiting on sweep.
1046 lock(&work.sweepWaiters.lock)
1048 injectglist(&work.sweepWaiters.list)
1049 unlock(&work.sweepWaiters.lock)
1051 // Increment the scavenge generation now.
1053 // This moment represents peak heap in use because we're
1054 // about to start sweeping.
1055 mheap_.pages.scav.index.nextGen()
1057 // Release the CPU limiter.
1058 gcCPULimiter.finishGCTransition(now)
1060 // Finish the current heap profiling cycle and start a new
1061 // heap profiling cycle. We do this before starting the world
1062 // so events don't leak into the wrong cycle.
1065 // There may be stale spans in mcaches that need to be swept.
1066 // Those aren't tracked in any sweep lists, so we need to
1067 // count them against sweep completion until we ensure all
1068 // those spans have been forced out.
1070 // If gcSweep fully swept the heap (for example if the sweep
1071 // is not concurrent due to a GODEBUG setting), then we expect
1072 // the sweepLocker to be invalid, since sweeping is done.
1074 // N.B. Below we might duplicate some work from gcSweep; this is
1075 // fine as all that work is idempotent within a GC cycle, and
1076 // we're still holding worldsema so a new cycle can't start.
1077 sl := sweep.active.begin()
1078 if !stwSwept && !sl.valid {
1079 throw("failed to set sweep barrier")
1080 } else if stwSwept && sl.valid {
1081 throw("non-concurrent sweep failed to drain all sweep queues")
1084 systemstack(func() { startTheWorldWithSema() })
1086 // Flush the heap profile so we can start a new cycle next GC.
1087 // This is relatively expensive, so we don't do it with the
1091 // Prepare workbufs for freeing by the sweeper. We do this
1092 // asynchronously because it can take non-trivial time.
1093 prepareFreeWorkbufs()
1095 // Free stack spans. This must be done between GC cycles.
1096 systemstack(freeStackSpans)
1098 // Ensure all mcaches are flushed. Each P will flush its own
1099 // mcache before allocating, but idle Ps may not. Since this
1100 // is necessary to sweep all spans, we need to ensure all
1101 // mcaches are flushed before we start the next GC cycle.
1103 // While we're here, flush the page cache for idle Ps to avoid
1104 // having pages get stuck on them. These pages are hidden from
1105 // the scavenger, so in small idle heaps a significant amount
1106 // of additional memory might be held onto.
1108 // Also, flush the pinner cache, to avoid leaking that memory
1110 forEachP(waitReasonFlushProcCaches, func(pp *p) {
1111 pp.mcache.prepareForSweep()
1112 if pp.status == _Pidle {
1113 systemstack(func() {
1115 pp.pcache.flush(&mheap_.pages)
1116 unlock(&mheap_.lock)
1119 pp.pinnerCache = nil
1122 // Now that we've swept stale spans in mcaches, they don't
1123 // count against unswept spans.
1125 // Note: this sweepLocker may not be valid if sweeping had
1126 // already completed during the STW. See the corresponding
1127 // begin() call that produced sl.
1128 sweep.active.end(sl)
1131 // Print gctrace before dropping worldsema. As soon as we drop
1132 // worldsema another cycle could start and smash the stats
1133 // we're trying to print.
1134 if debug.gctrace > 0 {
1135 util := int(memstats.gc_cpu_fraction * 100)
1139 print("gc ", memstats.numgc,
1140 " @", string(itoaDiv(sbuf[:], uint64(work.tSweepTerm-runtimeInitTime)/1e6, 3)), "s ",
1142 prev := work.tSweepTerm
1143 for i, ns := range []int64{work.tMark, work.tMarkTerm, work.tEnd} {
1147 print(string(fmtNSAsMS(sbuf[:], uint64(ns-prev))))
1150 print(" ms clock, ")
1151 for i, ns := range []int64{
1152 int64(work.stwprocs) * (work.tMark - work.tSweepTerm),
1153 gcController.assistTime.Load(),
1154 gcController.dedicatedMarkTime.Load() + gcController.fractionalMarkTime.Load(),
1155 gcController.idleMarkTime.Load(),
1158 if i == 2 || i == 3 {
1159 // Separate mark time components with /.
1164 print(string(fmtNSAsMS(sbuf[:], uint64(ns))))
1167 work.heap0>>20, "->", work.heap1>>20, "->", work.heap2>>20, " MB, ",
1168 gcController.lastHeapGoal>>20, " MB goal, ",
1169 gcController.lastStackScan.Load()>>20, " MB stacks, ",
1170 gcController.globalsScan.Load()>>20, " MB globals, ",
1171 work.maxprocs, " P")
1172 if work.userForced {
1179 // Set any arena chunks that were deferred to fault.
1180 lock(&userArenaState.lock)
1181 faultList := userArenaState.fault
1182 userArenaState.fault = nil
1183 unlock(&userArenaState.lock)
1184 for _, lc := range faultList {
1185 lc.mspan.setUserArenaChunkToFault()
1188 // Enable huge pages on some metadata if we cross a heap threshold.
1189 if gcController.heapGoal() > minHeapForMetadataHugePages {
1190 mheap_.enableMetadataHugePages()
1193 semrelease(&worldsema)
1195 // Careful: another GC cycle may start now.
1200 // now that gc is done, kick off finalizer thread if needed
1201 if !concurrentSweep {
1202 // give the queued finalizers, if any, a chance to run
1207 // gcBgMarkStartWorkers prepares background mark worker goroutines. These
1208 // goroutines will not run until the mark phase, but they must be started while
1209 // the work is not stopped and from a regular G stack. The caller must hold
1211 func gcBgMarkStartWorkers() {
1212 // Background marking is performed by per-P G's. Ensure that each P has
1213 // a background GC G.
1215 // Worker Gs don't exit if gomaxprocs is reduced. If it is raised
1216 // again, we can reuse the old workers; no need to create new workers.
1217 for gcBgMarkWorkerCount < gomaxprocs {
1220 notetsleepg(&work.bgMarkReady, -1)
1221 noteclear(&work.bgMarkReady)
1222 // The worker is now guaranteed to be added to the pool before
1223 // its P's next findRunnableGCWorker.
1225 gcBgMarkWorkerCount++
1229 // gcBgMarkPrepare sets up state for background marking.
1230 // Mutator assists must not yet be enabled.
1231 func gcBgMarkPrepare() {
1232 // Background marking will stop when the work queues are empty
1233 // and there are no more workers (note that, since this is
1234 // concurrent, this may be a transient state, but mark
1235 // termination will clean it up). Between background workers
1236 // and assists, we don't really know how many workers there
1237 // will be, so we pretend to have an arbitrarily large number
1238 // of workers, almost all of which are "waiting". While a
1239 // worker is working it decrements nwait. If nproc == nwait,
1240 // there are no workers.
1241 work.nproc = ^uint32(0)
1242 work.nwait = ^uint32(0)
1245 // gcBgMarkWorkerNode is an entry in the gcBgMarkWorkerPool. It points to a single
1246 // gcBgMarkWorker goroutine.
1247 type gcBgMarkWorkerNode struct {
1248 // Unused workers are managed in a lock-free stack. This field must be first.
1251 // The g of this worker.
1254 // Release this m on park. This is used to communicate with the unlock
1255 // function, which cannot access the G's stack. It is unused outside of
1256 // gcBgMarkWorker().
1260 func gcBgMarkWorker() {
1263 // We pass node to a gopark unlock function, so it can't be on
1264 // the stack (see gopark). Prevent deadlock from recursively
1265 // starting GC by disabling preemption.
1266 gp.m.preemptoff = "GC worker init"
1267 node := new(gcBgMarkWorkerNode)
1268 gp.m.preemptoff = ""
1272 node.m.set(acquirem())
1273 notewakeup(&work.bgMarkReady)
1274 // After this point, the background mark worker is generally scheduled
1275 // cooperatively by gcController.findRunnableGCWorker. While performing
1276 // work on the P, preemption is disabled because we are working on
1277 // P-local work buffers. When the preempt flag is set, this puts itself
1278 // into _Gwaiting to be woken up by gcController.findRunnableGCWorker
1279 // at the appropriate time.
1281 // When preemption is enabled (e.g., while in gcMarkDone), this worker
1282 // may be preempted and schedule as a _Grunnable G from a runq. That is
1283 // fine; it will eventually gopark again for further scheduling via
1284 // findRunnableGCWorker.
1286 // Since we disable preemption before notifying bgMarkReady, we
1287 // guarantee that this G will be in the worker pool for the next
1288 // findRunnableGCWorker. This isn't strictly necessary, but it reduces
1289 // latency between _GCmark starting and the workers starting.
1292 // Go to sleep until woken by
1293 // gcController.findRunnableGCWorker.
1294 gopark(func(g *g, nodep unsafe.Pointer) bool {
1295 node := (*gcBgMarkWorkerNode)(nodep)
1297 if mp := node.m.ptr(); mp != nil {
1298 // The worker G is no longer running; release
1301 // N.B. it is _safe_ to release the M as soon
1302 // as we are no longer performing P-local mark
1305 // However, since we cooperatively stop work
1306 // when gp.preempt is set, if we releasem in
1307 // the loop then the following call to gopark
1308 // would immediately preempt the G. This is
1309 // also safe, but inefficient: the G must
1310 // schedule again only to enter gopark and park
1311 // again. Thus, we defer the release until
1312 // after parking the G.
1316 // Release this G to the pool.
1317 gcBgMarkWorkerPool.push(&node.node)
1318 // Note that at this point, the G may immediately be
1319 // rescheduled and may be running.
1321 }, unsafe.Pointer(node), waitReasonGCWorkerIdle, traceBlockSystemGoroutine, 0)
1323 // Preemption must not occur here, or another G might see
1324 // p.gcMarkWorkerMode.
1326 // Disable preemption so we can use the gcw. If the
1327 // scheduler wants to preempt us, we'll stop draining,
1328 // dispose the gcw, and then preempt.
1329 node.m.set(acquirem())
1330 pp := gp.m.p.ptr() // P can't change with preemption disabled.
1332 if gcBlackenEnabled == 0 {
1333 println("worker mode", pp.gcMarkWorkerMode)
1334 throw("gcBgMarkWorker: blackening not enabled")
1337 if pp.gcMarkWorkerMode == gcMarkWorkerNotWorker {
1338 throw("gcBgMarkWorker: mode not set")
1341 startTime := nanotime()
1342 pp.gcMarkWorkerStartTime = startTime
1343 var trackLimiterEvent bool
1344 if pp.gcMarkWorkerMode == gcMarkWorkerIdleMode {
1345 trackLimiterEvent = pp.limiterEvent.start(limiterEventIdleMarkWork, startTime)
1348 decnwait := atomic.Xadd(&work.nwait, -1)
1349 if decnwait == work.nproc {
1350 println("runtime: work.nwait=", decnwait, "work.nproc=", work.nproc)
1351 throw("work.nwait was > work.nproc")
1354 systemstack(func() {
1355 // Mark our goroutine preemptible so its stack
1356 // can be scanned. This lets two mark workers
1357 // scan each other (otherwise, they would
1358 // deadlock). We must not modify anything on
1359 // the G stack. However, stack shrinking is
1360 // disabled for mark workers, so it is safe to
1361 // read from the G stack.
1362 casGToWaiting(gp, _Grunning, waitReasonGCWorkerActive)
1363 switch pp.gcMarkWorkerMode {
1365 throw("gcBgMarkWorker: unexpected gcMarkWorkerMode")
1366 case gcMarkWorkerDedicatedMode:
1367 gcDrainMarkWorkerDedicated(&pp.gcw, true)
1369 // We were preempted. This is
1370 // a useful signal to kick
1371 // everything out of the run
1372 // queue so it can run
1374 if drainQ, n := runqdrain(pp); n > 0 {
1376 globrunqputbatch(&drainQ, int32(n))
1380 // Go back to draining, this time
1381 // without preemption.
1382 gcDrainMarkWorkerDedicated(&pp.gcw, false)
1383 case gcMarkWorkerFractionalMode:
1384 gcDrainMarkWorkerFractional(&pp.gcw)
1385 case gcMarkWorkerIdleMode:
1386 gcDrainMarkWorkerIdle(&pp.gcw)
1388 casgstatus(gp, _Gwaiting, _Grunning)
1391 // Account for time and mark us as stopped.
1393 duration := now - startTime
1394 gcController.markWorkerStop(pp.gcMarkWorkerMode, duration)
1395 if trackLimiterEvent {
1396 pp.limiterEvent.stop(limiterEventIdleMarkWork, now)
1398 if pp.gcMarkWorkerMode == gcMarkWorkerFractionalMode {
1399 atomic.Xaddint64(&pp.gcFractionalMarkTime, duration)
1402 // Was this the last worker and did we run out
1404 incnwait := atomic.Xadd(&work.nwait, +1)
1405 if incnwait > work.nproc {
1406 println("runtime: p.gcMarkWorkerMode=", pp.gcMarkWorkerMode,
1407 "work.nwait=", incnwait, "work.nproc=", work.nproc)
1408 throw("work.nwait > work.nproc")
1411 // We'll releasem after this point and thus this P may run
1412 // something else. We must clear the worker mode to avoid
1413 // attributing the mode to a different (non-worker) G in
1415 pp.gcMarkWorkerMode = gcMarkWorkerNotWorker
1417 // If this worker reached a background mark completion
1418 // point, signal the main GC goroutine.
1419 if incnwait == work.nproc && !gcMarkWorkAvailable(nil) {
1420 // We don't need the P-local buffers here, allow
1421 // preemption because we may schedule like a regular
1422 // goroutine in gcMarkDone (block on locks, etc).
1423 releasem(node.m.ptr())
1431 // gcMarkWorkAvailable reports whether executing a mark worker
1432 // on p is potentially useful. p may be nil, in which case it only
1433 // checks the global sources of work.
1434 func gcMarkWorkAvailable(p *p) bool {
1435 if p != nil && !p.gcw.empty() {
1438 if !work.full.empty() {
1439 return true // global work available
1441 if work.markrootNext < work.markrootJobs {
1442 return true // root scan work available
1447 // gcMark runs the mark (or, for concurrent GC, mark termination)
1448 // All gcWork caches must be empty.
1449 // STW is in effect at this point.
1450 func gcMark(startTime int64) {
1451 if debug.allocfreetrace > 0 {
1455 if gcphase != _GCmarktermination {
1456 throw("in gcMark expecting to see gcphase as _GCmarktermination")
1458 work.tstart = startTime
1460 // Check that there's no marking work remaining.
1461 if work.full != 0 || work.markrootNext < work.markrootJobs {
1462 print("runtime: full=", hex(work.full), " next=", work.markrootNext, " jobs=", work.markrootJobs, " nDataRoots=", work.nDataRoots, " nBSSRoots=", work.nBSSRoots, " nSpanRoots=", work.nSpanRoots, " nStackRoots=", work.nStackRoots, "\n")
1463 panic("non-empty mark queue after concurrent mark")
1466 if debug.gccheckmark > 0 {
1467 // This is expensive when there's a large number of
1468 // Gs, so only do it if checkmark is also enabled.
1472 // Drop allg snapshot. allgs may have grown, in which case
1473 // this is the only reference to the old backing store and
1474 // there's no need to keep it around.
1475 work.stackRoots = nil
1477 // Clear out buffers and double-check that all gcWork caches
1478 // are empty. This should be ensured by gcMarkDone before we
1479 // enter mark termination.
1481 // TODO: We could clear out buffers just before mark if this
1482 // has a non-negligible impact on STW time.
1483 for _, p := range allp {
1484 // The write barrier may have buffered pointers since
1485 // the gcMarkDone barrier. However, since the barrier
1486 // ensured all reachable objects were marked, all of
1487 // these must be pointers to black objects. Hence we
1488 // can just discard the write barrier buffer.
1489 if debug.gccheckmark > 0 {
1490 // For debugging, flush the buffer and make
1491 // sure it really was all marked.
1500 print("runtime: P ", p.id, " flushedWork ", gcw.flushedWork)
1501 if gcw.wbuf1 == nil {
1502 print(" wbuf1=<nil>")
1504 print(" wbuf1.n=", gcw.wbuf1.nobj)
1506 if gcw.wbuf2 == nil {
1507 print(" wbuf2=<nil>")
1509 print(" wbuf2.n=", gcw.wbuf2.nobj)
1512 throw("P has cached GC work at end of mark termination")
1514 // There may still be cached empty buffers, which we
1515 // need to flush since we're going to free them. Also,
1516 // there may be non-zero stats because we allocated
1517 // black after the gcMarkDone barrier.
1521 // Flush scanAlloc from each mcache since we're about to modify
1522 // heapScan directly. If we were to flush this later, then scanAlloc
1523 // might have incorrect information.
1525 // Note that it's not important to retain this information; we know
1526 // exactly what heapScan is at this point via scanWork.
1527 for _, p := range allp {
1535 // Reset controller state.
1536 gcController.resetLive(work.bytesMarked)
1539 // gcSweep must be called on the system stack because it acquires the heap
1540 // lock. See mheap for details.
1542 // Returns true if the heap was fully swept by this function.
1544 // The world must be stopped.
1547 func gcSweep(mode gcMode) bool {
1548 assertWorldStopped()
1550 if gcphase != _GCoff {
1551 throw("gcSweep being done but phase is not GCoff")
1555 mheap_.sweepgen += 2
1556 sweep.active.reset()
1557 mheap_.pagesSwept.Store(0)
1558 mheap_.sweepArenas = mheap_.allArenas
1559 mheap_.reclaimIndex.Store(0)
1560 mheap_.reclaimCredit.Store(0)
1561 unlock(&mheap_.lock)
1563 sweep.centralIndex.clear()
1565 if !concurrentSweep || mode == gcForceBlockMode {
1566 // Special case synchronous sweep.
1567 // Record that no proportional sweeping has to happen.
1569 mheap_.sweepPagesPerByte = 0
1570 unlock(&mheap_.lock)
1571 // Flush all mcaches.
1572 for _, pp := range allp {
1573 pp.mcache.prepareForSweep()
1575 // Sweep all spans eagerly.
1576 for sweepone() != ^uintptr(0) {
1578 // Free workbufs eagerly.
1579 prepareFreeWorkbufs()
1580 for freeSomeWbufs(false) {
1582 // All "free" events for this mark/sweep cycle have
1583 // now happened, so we can make this profile cycle
1584 // available immediately.
1590 // Background sweep.
1593 sweep.parked = false
1594 ready(sweep.g, 0, true)
1600 // gcResetMarkState resets global state prior to marking (concurrent
1601 // or STW) and resets the stack scan state of all Gs.
1603 // This is safe to do without the world stopped because any Gs created
1604 // during or after this will start out in the reset state.
1606 // gcResetMarkState must be called on the system stack because it acquires
1607 // the heap lock. See mheap for details.
1610 func gcResetMarkState() {
1611 // This may be called during a concurrent phase, so lock to make sure
1612 // allgs doesn't change.
1613 forEachG(func(gp *g) {
1614 gp.gcscandone = false // set to true in gcphasework
1615 gp.gcAssistBytes = 0
1618 // Clear page marks. This is just 1MB per 64GB of heap, so the
1619 // time here is pretty trivial.
1621 arenas := mheap_.allArenas
1622 unlock(&mheap_.lock)
1623 for _, ai := range arenas {
1624 ha := mheap_.arenas[ai.l1()][ai.l2()]
1625 for i := range ha.pageMarks {
1630 work.bytesMarked = 0
1631 work.initialHeapLive = gcController.heapLive.Load()
1634 // Hooks for other packages
1636 var poolcleanup func()
1637 var boringCaches []unsafe.Pointer // for crypto/internal/boring
1639 //go:linkname sync_runtime_registerPoolCleanup sync.runtime_registerPoolCleanup
1640 func sync_runtime_registerPoolCleanup(f func()) {
1644 //go:linkname boring_registerCache crypto/internal/boring/bcache.registerCache
1645 func boring_registerCache(p unsafe.Pointer) {
1646 boringCaches = append(boringCaches, p)
1651 if poolcleanup != nil {
1655 // clear boringcrypto caches
1656 for _, p := range boringCaches {
1657 atomicstorep(p, nil)
1660 // Clear central sudog cache.
1661 // Leave per-P caches alone, they have strictly bounded size.
1662 // Disconnect cached list before dropping it on the floor,
1663 // so that a dangling ref to one entry does not pin all of them.
1664 lock(&sched.sudoglock)
1665 var sg, sgnext *sudog
1666 for sg = sched.sudogcache; sg != nil; sg = sgnext {
1670 sched.sudogcache = nil
1671 unlock(&sched.sudoglock)
1673 // Clear central defer pool.
1674 // Leave per-P pools alone, they have strictly bounded size.
1675 lock(&sched.deferlock)
1676 // disconnect cached list before dropping it on the floor,
1677 // so that a dangling ref to one entry does not pin all of them.
1678 var d, dlink *_defer
1679 for d = sched.deferpool; d != nil; d = dlink {
1683 sched.deferpool = nil
1684 unlock(&sched.deferlock)
1689 // itoaDiv formats val/(10**dec) into buf.
1690 func itoaDiv(buf []byte, val uint64, dec int) []byte {
1693 for val >= 10 || i >= idec {
1694 buf[i] = byte(val%10 + '0')
1702 buf[i] = byte(val + '0')
1706 // fmtNSAsMS nicely formats ns nanoseconds as milliseconds.
1707 func fmtNSAsMS(buf []byte, ns uint64) []byte {
1709 // Format as whole milliseconds.
1710 return itoaDiv(buf, ns/1e6, 0)
1712 // Format two digits of precision, with at most three decimal places.
1723 return itoaDiv(buf, x, dec)
1726 // Helpers for testing GC.
1728 // gcTestMoveStackOnNextCall causes the stack to be moved on a call
1729 // immediately following the call to this. It may not work correctly
1730 // if any other work appears after this call (such as returning).
1731 // Typically the following call should be marked go:noinline so it
1732 // performs a stack check.
1734 // In rare cases this may not cause the stack to move, specifically if
1735 // there's a preemption between this call and the next.
1736 func gcTestMoveStackOnNextCall() {
1738 gp.stackguard0 = stackForceMove
1741 // gcTestIsReachable performs a GC and returns a bit set where bit i
1742 // is set if ptrs[i] is reachable.
1743 func gcTestIsReachable(ptrs ...unsafe.Pointer) (mask uint64) {
1744 // This takes the pointers as unsafe.Pointers in order to keep
1745 // them live long enough for us to attach specials. After
1746 // that, we drop our references to them.
1749 panic("too many pointers for uint64 mask")
1752 // Block GC while we attach specials and drop our references
1753 // to ptrs. Otherwise, if a GC is in progress, it could mark
1754 // them reachable via this function before we have a chance to
1758 // Create reachability specials for ptrs.
1759 specials := make([]*specialReachable, len(ptrs))
1760 for i, p := range ptrs {
1761 lock(&mheap_.speciallock)
1762 s := (*specialReachable)(mheap_.specialReachableAlloc.alloc())
1763 unlock(&mheap_.speciallock)
1764 s.special.kind = _KindSpecialReachable
1765 if !addspecial(p, &s.special) {
1766 throw("already have a reachable special (duplicate pointer?)")
1769 // Make sure we don't retain ptrs.
1775 // Force a full GC and sweep.
1778 // Process specials.
1779 for i, s := range specials {
1782 println("runtime: object", i, "was not swept")
1783 throw("IsReachable failed")
1788 lock(&mheap_.speciallock)
1789 mheap_.specialReachableAlloc.free(unsafe.Pointer(s))
1790 unlock(&mheap_.speciallock)
1796 // gcTestPointerClass returns the category of what p points to, one of:
1797 // "heap", "stack", "data", "bss", "other". This is useful for checking
1798 // that a test is doing what it's intended to do.
1800 // This is nosplit simply to avoid extra pointer shuffling that may
1801 // complicate a test.
1804 func gcTestPointerClass(p unsafe.Pointer) string {
1805 p2 := uintptr(noescape(p))
1807 if gp.stack.lo <= p2 && p2 < gp.stack.hi {
1810 if base, _, _ := findObject(p2, 0, 0); base != 0 {
1813 for _, datap := range activeModules() {
1814 if datap.data <= p2 && p2 < datap.edata || datap.noptrdata <= p2 && p2 < datap.enoptrdata {
1817 if datap.bss <= p2 && p2 < datap.ebss || datap.noptrbss <= p2 && p2 <= datap.enoptrbss {